‘For the rational study of the law the black letter man may be the man of the present, but the man of the future is the man of statistics. . .’
Oliver Wendell Holmes, 1897Footnote 1
1. Introduction
In a recent book, Lee Epstein and Andrew D. Martin dryly remark that ‘[t]o claim that research based on data – that is, empirical work – has infiltrated the legal community borders on the boring’.Footnote 2 What is less boring, however, is the question as to the wider implications for international law (IL) of this veritable boom of empirical scholarship predicted more than a century ago by Wendell Holmes. This turn to empirical analysis of law is for some a sign of progress of legal science – even its ability to find better solutions to real life international problems.Footnote 3 Others call for caution and even show hostility to these forms of IL studies.Footnote 4 At a first glance, this at times heated debate invokes all the classic dichotomies of the epistemology of legal science: Sollen v. Sein, doctrinal v. empirical, normative v. descriptive, norms v. facts, etc. Nevertheless, the current scholarly debate seems to have considerable problems keeping a clear focus on the underlying epistemological framework in which each of the positions is inscribed and how they possibly relate or do not relate to one another.
With a view to the question of the future of international legal methodology, the real question therefore very much remains: what are the implications of this manifest growth in empirical approaches – what are we to make of this in terms of the philosophy of international legal science, of the social understanding of IL, and, not the least, the place of doctrinal scholarship after the alleged empirical Wende. What is needed to really begin to answer these questions, we submit, is not yet another partisan suggestion for the right position on this controversial issue, but rather an attempt at making intelligible both the oppositions and the possibilities of synthesis between normative and empirical approaches to law.Footnote 5 In this article we explore the implications of the alleged empirical turn from the point of view of legal scholarship as traditionally assumed to be doctrinal scholarship interested in the normative order of law as somewhat of a coherent matrix. We therefore propose a general conceptual model with regard to the interface of doctrinal law and empirical studies of lawFootnote 6 and their consequences for the future of legal scholarship. Based on our assessment and rational reconstruction of the arguments and positions assumed in the literature, we sketch out three basic ideal types in terms of the epistemological understanding of the interface of law and empirical studies: toleration, synthesis and replacement.
By toleration we refer to the peaceful but asymmetric co-existence between doctrinal law and empirical studies of law but with doctrinal law in the commanding position. This situation is founded on the premise of a relative supremacy of doctrinal law over empirical studies of law and a corresponding inside/outside dichotomy where doctrinal law is the unavoidable and essential insider. The consequence for the outside discipline is that it has to accept the role as a sort of ‘science auxiliaire’Footnote 7 or ‘science supplémentaire’. In some versions of toleration there is an asymmetric conceptual dependency: empirical studies of law presuppose the validity of doctrinal law as science – but not vice versa.Footnote 8 This implies that there is a conceptual necessity of doctrinal law as it provides the premise for a sociology of law or similar outside perspectives. From the point of view of international doctrinal legal scholarship, the empirical turn thus conceived becomes the story of an unfortunate displacement of scholarly energy and emphasis potentially turning the once auxiliary discipline into the main discipline. Accordingly, the doctrinal perspective should remain foundational and supreme and genuine research on law is in that sense effectively mono-disciplinary.
By synthesis, or in some cases necessary synthesis, we refer to the peaceful co-existence between doctrinal law and empirical studies of law based on the mutual recognition of the key value of the respective disciplines for understanding law in its totality. This position is marked by the notion that doctrinal studies can be enlightened by empirical studies and vice versa. This more ‘weak programme’ assumes that both are to benefit but it is, as we will discuss below, often silent on the precise epistemological premise for such collaboration. From the point of view of international doctrinal scholarship, the empirical turn thus conceived becomes the story of merely a change in relative emphasis between law and empirical context but ultimately the notion of synthesis remains unchallenged. Its methodological trademark is a form of transdisciplinarity.
Replacement describes in its strongest version the position that conceives synthesis or even co-existence to be impossible as empirical studies of law erases and/or replaces doctrinal law as the premise for understanding law. This situation is marked by a clash between perspectives and approaches. From the point of view of doctrinal law, the advocacy for the turn to empirics has the character of a ‘hostile take-over’. It constitutes a foundational critique, which in essence denies that doctrinal law is a sound science. In some cases, it is a form of error theory as empirical studies of IL not only make the claim that these forms of knowledge do not exist, but also explain why agents have nevertheless come to (wrongly) believe such knowledge exists.Footnote 9 From the point of view of international legal scholarship, the empirical turn thus conceived becomes the story of a fundamental challenge to the existence of doctrinal law as scientific paradigm. In its strongest version, it thus defies doctrinal scholarship as a scientific possibility. It, however, also comes in a less radical variation where empirical studies of IL are conducted in what effectively is a separate world from doctrinal law – it has different audiences and scientific premises but it does not engage directly with doctrinal law. In this weak version, it is hardly a competing perspective but a separate one, which only shares with doctrinal law the common focus on the field of IL as object of inquiry.
Using these ideal-types for structuring the debate introduces of course a particular way of framing this question, and one that emphasizes the epistemological level. Most previous scholarship has instead focused on empirical properties in terms of the relationship between law and politics.Footnote 10 But the resulting debates leave us with unsatisfying answers and antagonistic view-points, ranging from hostility to claims of interdisciplinarity, multidisciplinarity and even transdisciplinarity or convergence. That is the reason we use a markedly different strategy. Starting inductively on the basis of the input from previous debates our approach is to exhaust the logical space of possibilities by introducing our three broad ideal-types. Also, we differ from some main camps in the law-politics debate as we take very seriously the épistème of doctrinal law. We, however, do not do this with the aim of presenting a partisan position in favour of doctrinal law. Rather, we introduce doctrinal law as our comparator in order to emphasize how empirical legal scholarship looks from the point of view of the dominant paradigm of legal research. This also means that when we refer to empirical scholarship we do so with a starting-point in the well-known and recognized distinction in legal philosophy (and philosophy of science more generally) between is and ought, as one way of drawing the boundary between objects that can be empirical and objects that cannot. As we have elaborated elsewhere, there is an empirical element in doctrinal law, notably in terms of the place of case law. The treatment of this material, however, is not characteristic of an empirical science. Rather, doctrinal law is based on an axiological presumption of validity that differs radically from empirical validity.Footnote 11 Doctrinal law is therefore, in our view, not an empirical science.
The article is structured around the three outlined ideal-typical representations of the interrelationship between law and empirical approaches to law and their consequences for the science of IL. In each of the three sections, we provide examples of the respective positions which we consider representative of the outlined ideal-typical positions; the examples are in other words not as such exhaustive but used mainly to further refine the ideal-typical positions. In each case, we first lay out the epistemological premises for each of the positions and then analyze examples of both historical and current scholarship that is representative of the position. Against this background we conclude by a more general analysis of the consequences of the empirical turn for international legal scholarship. For reasons of space, the following presentation of examples of each position is necessarily highly selective. Instead of a comprehensive listing, we have prioritized, giving examples sufficient to articulate the principled position and a sense of some key variations in actual literature.
2. Interrelations of doctrinal law and empirical legal studies: three positions
2.1. Toleration
The first approach to the ongoing rise in empirical studies of IL law is also the one most commonly adopted in doctrinal legal scholarship. We describe it as toleration in the sense that proponents accept the presence and even the legitimacy of empirical studies of (international) law per se, but they do so only somewhat reluctantly and while emphasizing the subordinate or auxiliary character of such studies vis-á-vis the mother discipline which remains doctrinal legal science. From this point of view, then, there is nothing inherently wrong or epistemologically problematic in engaging in empirical approaches to IL. However, the current boom in such approaches is seen more generally as a regrettable development because it takes time and resources away from the primary issues which continue to require traditional doctrinal approaches.
Armstrong expressed the gist of this approach of toleration in his rhetorical question: ‘Political science has discovered the European Court of Justice (ECJ). But has it discovered law?’Footnote 12 The position is naturally closely associated with the classic positivist theories of doctrinal legal science articulated first of all by Hans Kelsen but also, in significant respects, by H.L.A. Hart.Footnote 13 Simplifying somewhat, proponents of toleration base their reserved attitude towards empirical studies on two main arguments, which although very often found in tandem are logically distinct and can be adopted independently.
2.1.1. The uniqueness of law as normative science
First of all, they maintain that a purely empirical approach cannot adequately capture the true character of the (international) legal field in its entirety; that it will necessarily leave a quintessential aspect of that field unexplained. It bases this claim on the assumption of a categorical divide between Sein and Sollen, between facts and norms, between the descriptive and the normative. Observing that law consists essentially of legal rules which are normative phenomena (distributing legal rights and obligations) proponents of toleration infer that law as such necessarily remains categorically impervious to empirical studies.
Kelsen notoriously maintains that empirical disciplines (or, in his terminology: natural science) can only study the Sein of law and that it can therefore never (or only on pain of committing the naturalistic fallacy) justify claims about the Sollen of law:
Nobody can deny that the statement: “something is” – that is, the statement by which an existent fact is described – is fundamentally different from the statement: “something ought to be” – which is the statement by which a norm is described. Nobody can assert that from the statement that something is, follows a statement that something ought to be, or vice versa.Footnote 14
Correspondingly, although from a slightly different angle, Hart maintains that empirical approaches can only observe the so-called external aspect of social rules, notably the regularities of group behaviour, and that it can never have access to the internal aspect which therefore remains the business of doctrinal study of law. To be sure, Hart's theory comes somewhat closer to adopting an empirical approach than does Kelsen's. Hart famously describes The Concept of Law ‘as an essay in descriptive sociology’, and his theory is often referred to as social fact-based positivism because, unlike Kelsen who places a hypohetically presupposed Grundnorm at the foundation of his theory, Hart takes a social practice, i.e., the practice of law-applying authorities, to be the defining element of his rule of recognition. However, the difference between Hart and Kelsen on this issue can easily be overstated. First, Kelsen, in his own way, by no means ignores social practices. On the contrary, such practices manifestly provide the necessary condition for the identification of the relevant Grundnorm on his account.Footnote 15 Second, the identification of the rule of recognition in any legal system may indeed be a complicated empirical matter according to Hart.Footnote 16 However, once that particular rule has been positively identifiedFootnote 17 , law-ascertainment remains a question of engaging in inter-normative reasoning processes between this presupposed starting point and any primary legal rules asserted to be part of a given legal system.Footnote 18 This kind of reasoning may by different doctrinal scholars be practiced as either vulgar or sophisticated formalism Footnote 19 but either way it remains an exercise which is categorically different from anything found in the empirical disciplines. To both Kelsen and Hart and their present-day followers in doctrinal IL scholarship, empirical approaches are therefore to be applied only at the axiomatic level. Hence the perception among proponents of toleration of the continued distinctiveness or purity of doctrinal studies of IL, and the corresponding blanket rejection of any more ambitious imperial claims made by proponents of empirical approaches to the field.
2.1.2. The necessity of doctrinal law for empirical studies of law
The second main argument applied by at least some proponents of toleration is more radical. It holds that empirical scholars’ imperial claims are misguided not only because the empirical studies cannot exhaust the field of IL. Their imperial claims are misguided at an even more fundamental level, because they fail to appreciate the asymmetric interrelation between traditional normative doctrinal scholarship and empirical studies of law, which rightly understood renders the latter conceptually and epistemologically dependant on the former. For the legal empirical scholar to even begin studying their own external domain in the legal field, i.e., the ‘is’ which lies beyond the legal doctrinal ‘ought’, so the argument goes, they shall necessarily have to presuppose the validity of the discipline which studies this ‘ought’, i.e., the traditional doctrinal science of (international) law. This view is found in Kelsen's work – from very early on in his critique of EhrlichFootnote 20 and, later, also in his critiques of both WeberFootnote 21 and RossFootnote 22 . Here, Kelsen argues that any specifically legal empirical discipline (legal sociology, legal psychology, etc.) necessarily has to presuppose the validity of a normative jurisprudence, i.e., of traditional doctrinal studies of law. To illustrate, Kelsen in his critique of Ehrlich:
[T]he separation of this legal sociology can only be achieved by using a concept which originates in a categorically different view-point than that of an explicative sociology, i.e. in normativist scholarship's concept of law. Legal sociology cannot determine . . . what law is, as it has to presuppose normativist scholarship's concept of law.Footnote 23
Arguments to this effect have been applied repeatedly in recent debates about the boom in empirical approaches and about the associated call for interdisciplinarity in scholarship on IL.Footnote 24
Thus, to recapitulate, the position of toleration towards empirical legal studies maintains that even though we may be observing an increase in the absolute and relative number of empirical studies of (international) law, the very notion of an empirical turn is misguided in two ways. First, it is wrong on pragmatic grounds to turn exclusively to empirical studies, since this implies leaving highly important aspects of (international) law unexplained. Second, thus turning to empirical studies is impossible for principled epistemological reasons, simply because the possibility of such studies relies on the existence of doctrinal studies of IL. Or, more precisely: the former constitutes the possibility conditions of the latter. At best, then, instead of a celebration of the current development as an empirical turn we get from this point of view a picture of cautious, asymmetric toleration. Although it cannot generally be excluded that empirical studies may also provide interesting knowledge on matters related to (though not strictly on) law, they remain secondary and wholly contingent upon traditional doctrinal scholarship.
2.2. Synthesis: bridging law and empirical studies
The mutual recognition of and simultaneous engagement in both doctrinal legal science and variations of empirical studies of law constitutes the position we refer to as synthesis. As the other ideal-typical representations which we use for this article, the position can be differentiated into more specific positions but the defining feature is relatively peaceful co-existence based on the mutual recognition of the key value of each of the respective disciplines for understanding law in its totality. In some instances, there is even a sense of necessary synthesis, that is, there is a need of both perspectives: a study of the norms and a study of the making of the norms and the norms in action. A comprehensive understanding of law in other words, needs both. In most cases, however, the inherent epistemological tension between law and empirical approaches is solved by silence – or only scant reference to the basic scientific problems the approach entails, and which Kelsen found categorical and which have made scholars whom we have placed in the position of toleration call for a return to formalism. This is a position that is largely similar to what is found in so-called socio-legal studies where researchers, typically lawyers, combine doctrinal analysis with a number of empirical approaches.Footnote 25 With regard to IL, we will illustrate the position with first a brief discussion of the work of Joseph Weiler and Martti Koskenniemi who both, in different ways, however, have very successfully contextualized IL in their analyses. American New Legal Realism suggests a different kind of synthesis. A key proponent of this movement, Gregory Shaffer, argues that the approach is founded in pragmatism but, as we will discuss, it is a particular kind of pragmatism that seems to accept the foundationalism of doctrinal law in combination with the anti-foundationalism derived from pragmatism. These examples are of course not an exhaustive depiction of the position of synthesis but only a set of emblematic examples of scholarship within that position.
2.2.1. International law in context
In Weiler's classic analysis of the development of European law and the corresponding transformation of Europe, he only makes brief remarks on theory and methodology.Footnote 26 What Weiler does reveal is, however, interesting for the purpose of this article. Writing in response to Martin Shapiro, who at the point of publishing had criticized the use of the, then, novel notion of European constitutional law as a juridical construct leaving out the necessary political component for making constitutionalism intelligible,Footnote 27 Weiler sets out to ‘. . .analyse the Community constitutional order with particular regard to its living political matrix; the interactions between norms and norm-making, constitution and institutions, principles and practices, and the Court of Justice and the political organs. . .’.Footnote 28 But he adds: ‘. . .even though I shall look at relationships of legal structure and political process, at law and power, my approach is hardly one of Law in Context. . .’.Footnote 29 Indeed, and not without some self-irony, he depicts his approach as being in the tradition of the ‘pure theory of law’. Weiler justifies this (re)turn to the ‘inside’ of law through the fact that arguably a great part of the transformation of European law results from the ‘internal dynamics of the system itself, almost if it were insulated from those “external” aspects’.Footnote 30 The latter claim is inspired by scholarship in the area of systems theory, which claims that internal workings of the legal system to a large extent explain its evolution in terms of self-reference.
The result of the approach devised by Weiler is a seminal piece explaining, largely via developments at the intra-legal level but always contextualized more broadly to society and politics, how Europe has transformed by changes in EC/EU law. Ultimately it is a more structural analysis that is different from the Kelsenian expectations evoked by Weiler.Footnote 31 This successful attempt at synthesis is followed up in a series of later publications where particularly constitutionalism and constitutionalization are submitted to an analysis which is both legal-political historical and notably intellectual-historical. While always being the lawyer with a sharp eye for legal developments, Weiler's analyses nevertheless underlines how European legal integration studies are deeply inter-disciplinary.Footnote 32 They not only produce synthesis – they need synthesis in the first place according to Weiler. Weiler basically suggests a normative-descriptive bridge that produces the perhaps best ever analysis of European integration by law, but also leaves an open epistemological challenge to the most foundational ideas of basic science. One possible answer is that Weiler's analysis is more empirical than its self-presentation and indeed treats law from an outsider perspective in line with what we have elsewhere outlined as the European New Legal Realism.Footnote 33
Another author who we broadly position under the sub-rubric of ‘international law in context’ is Martti Koskenniemi, who almost single-handedly has spurred the current interest in more historical studies of IL. Like Weiler, Koskenniemmi, at least in his more recent scholarship on the history of IL, does not bore the reader with questions of epistemology. Of course, his earlier writings were theoretically highly ambitious and sought explicitly to combine descriptive and normative approaches in the analysis of law.Footnote 34 Indeed, a major criticism presented in his more theoretical pieces was that the turn to doctrine or a more bureaucratic role of IL was not only out of sync with the (original) mission of IL, it was self-defeating for the discipline. In line with a claim found more generally in Critical Legal Studies, Koskenniemmi's work is however not about law and politics but about how IL is an expression of politics.Footnote 35 This does not render it interdisciplinary in a common fashion. On the contrary, it is a work for international lawyers, thinking about and practicing IL, according to the author. There is, in other words, an attempt at what we call necessary synthesis but from the perspective of the lawyer. It seems to follow from this that IL is a particularly applied social and human science that, regardless of its differentiation as a highly specialized field and the importance of the intellectual and conceptual developments of that field, needs to be analyzed as part of a broader approach. Like Weiler, the result is more of a structural analysis of law, in this case the structure of the legal argument. Koskenniemmi, however, differs from Weiler in his continuous ambition of reflecting on the different styles of the practitioner of IL and of the academic of IL – it is the latter who employs the full portfolio of the humanities and social science. This position creates a form of multi-episteme that rejects – or better, ignores – the classic distinction between ought and is and instead emphasizes the different repertoires or styles of analysis available.Footnote 36
Koskenniemi's rather unique position – highly theoretical but not with a conventional theoretical objective – is further articulated in his work on the history of international law (and lawyers). Particularly in the Gentle Civilizer of Nations,Footnote 37 he undertakes a large-scale analysis of key thinkers of IL in four major Western countries in this regard. It is not a social history of international law and lawyers. Instead it is more in line with the history of ideas, notably its more modern incarnation following Quentin Skinner. Although clearly inspired by Skinner, Koskenniemi is not making many concessions to history or other relevant disciplines, including sociology. He dryly notes: ‘Lawyers – especially those with an interdisciplinary interest – should bear in mind that the grass is not necessarily any greener in the adjoining field.’Footnote 38 While his earlier work is about legal structure and is inspired by the developments in post-structuralist philosophy of the period, his more recent work is more historical according to the author. Ultimately, however, the more historical work is also structural regardless of the many biographic and political details it includes. The difference from both his own previous work and other histories of IL is that Koskenniemi seeks deliberately to avoid grand conceptual or epochal abstractions, even if the subtitle of his book, ‘The Rise and Fall of International Law’ seems to indicate the latter.Footnote 39 Overall, the result is a more freely narrated history at the intersection of some kind of legal formalism and history. Yet, Koskenniemi does not offer much on his stance on legal formalism and its usage in combination with empirical historical studies.Footnote 40 Formalism is somehow assumed, but it is never a hindrance for creating a synthesis with historical sources and an implicit normative view on IL. In the latter, Koskenniemi is more radical than Weiler and cannot be reconstructed as a present-day realist. Instead, his epistemological grounding is a multi-episteme.
2.2.2. American New Legal Realism
A different take on the problem of synthesis between legal formalism and empirical studies of law is found in the American revival of legal realism, at least in the version propounded by Gregory Shaffer. Shaffer makes a relativist attempt at solving the problem inherent in the position of synthesis. He seems to accept the foundational epistemology of doctrinal law as conceived in legal positivism but is critical of the completeness of this approach for explaining law more generally and IL in particular.Footnote 41 The solution provided by, particularly, Shaffer is a return to pragmatism:
The pragmatists were anti-foundationalist thinkers who stressed the importance of empirical work, combined with experimental practice aimed at problem-solving in particular social contexts. Law, from a pragmatist perspective, is a particular means for creating social order and social welfare . . . Today, given the expansion of international law's scope and its greater enmeshment with national law, new opportunities arise for transnational problem-oriented, pragmatist thinking about it.Footnote 42
While this at first glance seems like a complete turn to social science and socio-legal engineering, American New Realism in Shaffer's version never entirely gives up on the conceptual guidance on ‘what is law’ provided by legal positivism. Epistemologically, they are not in opposition (see on replacement below) but rather mutually informative, complementary ways of understanding IL. This leads to somewhat of an inconsistent position, which is critical of legal formalism only under specific circumstances, that is, when it is under-determinate.Footnote 43 There is an attempt at a synthesis but it is seemingly based on a notion of supplementary or ‘para-tactical’ epistemologies of which the hierarchy or their mutual implications are never precisely spelled out. For this reason it might even be argued that it belongs, in principle, in the category of toleration – if it were not for the predominant role empirical research on IL plays in the scholarship actually carried out by Shaffer and likeminded researchers.
This is also where Shaffer's position differs from American New Legal Realism more generally. Other proponents, notably Elizabeth Mertz and Bryant Garth, have more in line with the Law and Society paradigm of being both empirical and programmatic argued for a far greater (if not complete) rule-scepticism, indeed a position which fits better with the replacement position that we turn to now.Footnote 44 This latter position thereby deviates from the pragmatism at the heart of Shaffer's position and suggests a more conventional social scientific paradigm of analysis.
2.3. Replacement
Many of the approaches we highlight within the position of synthesis use empirical material of some kind in the analysis of law, but with the exception of American New Legal Realism, none of them rely on conventional social scientific methods. Such methods are far more prevalent in the third position, which we name replacement. Replacement in its strong form represents the most radical challenge to traditional doctrinal approaches to IL scholarship. This approach takes the idea of an empirical turn seriously, implicitly referring to other celebrated academic turns like Kant's Copernican revolution and the so-called ‘linguistic turn’.Footnote 45 In these contexts, the idea of a turn connotes radical and irreversible scholarly reorientation; the perception that a previously predominant approach to a given field has, for epistemological reasons, become out-dated, even obsolete and should be or is being replaced with a categorically different theoretical approach. Dubbing the current development an empirical turn in scholarship on ILFootnote 46 taps into this broader meaning applying it specifically to the question of whether doctrinal approaches are to be disposed of on epistemological grounds and replaced by empirical approaches across the board. Unsurprisingly, this fundamental challenge has triggered hefty responses from doctrinal scholars arguing it is essentially a hostile take-over of law.Footnote 47 In the following, we first present the basic tenets of the underlying epistemology of naturalism and then, secondly, discuss a set of studies which to varying degrees rely on such premises.
2.3.1. The Analogy with replacement naturalism as an epistemological programme
The empirical turn in IL scholarship is often claimed on empirical grounds – there are simply more empirical studies of IL than previously. In our view it can, however, more theoretically – and convincingly – be described in analogy with and within the framework of, what has in the context of epistemology and philosophy of science been described as replacement naturalism or just naturalism.Footnote 48 Naturalism, as we use the term here, takes its cue from the work of W.V. Quine.Footnote 49 We, however, apply a broader interpretation of this term in order for it to be able to serve as a framework for conceptualizing the third position in our taxonomy. In particular, we should emphasize right from the outset that, unlike Quine, we do not identify naturalism exclusively as a matter for natural science but see it more generally as a matter for empirical science en bloc, be it natural or social science.Footnote 50 On this more general understanding, then, naturalism refers to a two-pronged philosophical theory, which combines a negative claim about the prospects of providing a priori philosophical justification for science, and a positive or constructive claim about what scholars should do instead.
As to the negative claim, naturalism proclaims the inevitable failure of any attempt at providing a priori philosophical justification of science and knowledge. Quine focused his full attention on the specific foundationalist programme suggested by logical positivism but the claim can be generalized to all attempts to provide first philosophical justifications for science as such. This leads to the second, positive, claim of naturalism, which is of key relevance in the present context of the alleged empirical turn in IL scholarship. Quine makes the further point that not only is the traditional epistemological project doomed to fail, it is also fundamentally misconceived because it aims to save science from scepticism. Rightly understood, the relevant task for epistemology is not to combat scepticism but rather to record and explain the actual existence of science as an empirical matter of fact: ‘But why all this creative reconstruction, all this make-believe? . . . Why not just see how this construction [of our picture of the world] really proceeds?’Footnote 51
Instead of trying in the abstract to justify given scientific propositions or entire disciplines, as being correct or true, the relevant epistemological task is instead one of describing empirically how, as a matter of fact, certain propositions/disciplines have come to be thus considered by the scientific community. That is, instead of reviewing the abstract logical relationship between propositions in a given body of scientific theory that may or may not justify it or some portion of it, naturalized epistemology focuses squarely upon the actual social-psychological relationship between these propositions and members of the scientific community taken individually and as a group. In other words, it observes the fact that a given scientific discipline exists and studies empirically the ways in which members of the scientific community have come to hold this body of beliefs as scientific truths.
Studying this question is a matter for the empirical sciences and not for any first philosophy. It seems more helpful, however, to rethink naturalism in much broader terms so as to encompass that whole vigorous empirical turn in which a long line of empirical disciplines – from neuroscience through evolutionary biology and behaviourist psychology to sociology of science – have come to the fore and triumphantly claimed (apparently each discipline for itself) to be ‘heir of the subject that used to be called philosophy’, to use Wittgenstein's phrase.Footnote 52 With a view to the bigger picture what is important is that these different and often competing theoretical schools can be said to subscribe to the two main naturalist tenets as described above: i) the failure of justificatory foundationalism; and ii) the view that this normative ‘armchair’ programme should be replaced by a descriptive empirical study.
2.3.2. The replacement thesis in international law
Turning then to the alleged empirical turn in present day IL scholarship, the parallels to replacement naturalism thus conceived are apparent. Among some proponents of empirical approaches we seem to find an ambition to not only do ‘additional valuable’ academic work but rather a more openly antagonistic agenda to actively replace doctrinal approaches to IL. No longer satisfied with providing an auxiliary to traditional approaches they claim, deep down, to be the (new) legal science. Thus, we find, at least implicitly, what would seem to be an analogue two-step theory in many proponents of empirical approaches: on the one hand, a negative claim about traditional attempts in the philosophy of law to provide a priori foundations for the doctrinal study of law; on the other hand, a positive or constructive claim that the existence of IL and its institutions should or can therefore only meaningfully be studied and explained empirically.
This description seems, at least superficially, to fit quite a portion of the empirical scholarship on IL which comprises the current boom ranging from political science, Law and Economics, Empirical Legal Studies, European New Legal Realism and a long series of studies in the field of sociology of IL and other disciplines. Although to a varying degree, these all have in common the subscription to the above two key tenets of replacement naturalism. But they differ, at certain points radically, on a number of issues, first of all on which empirical disciplines they each think that IL scholarship should turn to.
While this replacement-thesis is quite clear in principle it nevertheless causes some problems in practice when applied to actual empirical studies of law. The problem is simply that in much of the empirical work on IL one rarely finds explicit consent to the first assumption underlying the replacement thesis, i.e., the negative claim that doctrinal studies of IL is no longer feasible because epistemologically unsound. This makes it perfectly possible (and sometimes it is also the case) that empirical scholars are genuinely agnostic about this question; i.e., that they perceive themselves merely to be doing academic work that is valuable in its own right regardless of its relation to doctrinal scholarship on IL, and holding no particular views on the soundness of the latter. Others, however, undoubtedly have bigger fish to fry. They turn to empirics not merely as another interesting thing to do in the Academy. Rather, they turn to empirics because they consider doctrinal scholarship flawed on epistemological grounds, and conversely find empirical approaches to be the adequate approach(es) for making the international legal field intelligible.
2.3.3. Empirical studies of international law
To illustrate this tension between replacement as an ideal type and actual instances of the empirical turn, we will in the following briefly analyze four distinct empirical approaches to IL, which to varying degrees fits the replacement thesis. The probably best known empirical approach to IL is found in political science where there has been a surge of interest among political scientists to study IL in recent years. While European legal institutions – and to a lesser extent European law – was, for some time, a fertile ground for law in context, blending legal and political science analysis, the turn to IL by empirically-oriented political scientists is a more recent but larger phenomenon. This phenomenon has drawn a lot of attention and it has also been received with widespread caution or outright scepticism by doctrinal scholars who perceived this development as a badly concealed hostile take-over.Footnote 53 However, a recent publication illustrates how this need not be the case. The flagship American Journal of International Law published in 2012 a 50-page essay outlining the state of the field of political science research on ILFootnote 54 in the same issue as an article tellingly entitled ‘The Empirical Turn in International Legal Scholarship’.Footnote 55 In spite of using the term turn, which would seem to signal an approval of some version of the replacement thesis, Hafner-Burton, Victor and Lupu, on a close reading, in fact make a somewhat different claim, that of a certain parallelism between doctrinal and IL approaches. Strictly, when seen from the vantage point of doctrinal IL, many of the IR studies of IL could consequently be situated in the toleration category identified above.
Hafner-Burton, Victor and Lupu start from the observation that overall the political scientific enterprise with regard to IL is simply distinct from that of academics of public international law. This is in part due to the differences in audience: while public international law scholarship is mainly directed at practitioners and policy-makers, most IL work by political scientists is aimed at graduate training. Moreover, political scientists do not differentiate IL as a unique source of power and co-operation in the world, but rather as part of a larger amalgam of forces at play in global politics. And while legal scholars inherently focus on the contents of law and legal reasoning, political scientists work with the more ambiguous idea of institutions.Footnote 56 Considering these real differences in objects, objectives and audiences many empirical studies of IL in political science are effectively carried out in a separate world from doctrinal law which neither criticizes nor seeks to replace doctrinal approaches to IL. Institutionally, they exist in a parallel academic universe but this obviously does not exclude that they can have ideational or logical implications also in the doctrinal realm.
As an exception to this trend, the authors, however, suggest that a growing overlap is found between doctrinal lawyers and International Relations scholars with regard to questions of institutional design and impact of international legal institutions. In our scheme of analysis, this suggests a growing synthesis, what one IR Scholar, Kenneth Abbott, had gone as far as describing as a ‘joint discipline’.Footnote 57 The question remains, however, whether the alleged underlying differences in Erkenntnisinteresse can be mended by common interests in more abstracted legal-political ideas, for example constitutionalism or the rule of law. Again the literature is rather silent on the foundational questions raised by naturalism and the replacement thesis. There is neither a negative argument for the insufficiency of the legal doctrinal position, nor a full-blown attempt to replace existing scholarship. There is an attempted co-existence which, however, is mainly based on acceptance of different projects and the relative value each of them bring to broader questions. In that sense we are dealing with a literature which either seeks a more vague synthesis which, however, is different from the IL in context outlined above as it is explicitly based on political science analytical frameworks or it is simply carried out in an academic parallel universe.
A more direct onslaught on legal formalism, however, is found in some sociology of international law. An example is the work of Dezalay and Garth who explain the evolution of international commercial arbitration practically without reference to the legal contents of the field. Instead, they show how the struggle over the form of international commercial arbitration is fundamental. They explain this as a battle between different forms of expertise derived from competing global elites. But, they always stop short of using these structural insights for explaining the more ‘normative’ side of law and legal contents. The argument for doing so is that law – because of its normative shadow and the vested professional interests in depicting law in specific ways – is a real hindrance for a rigorous understanding of what law actually is. This is a deeply Bourdieusian argument, namely that it is methodologically necessary to rupture with the powerful sociolinguistic mechanisms of law, that is, the officializing and objectivizing discourses of law, in order to actually understand law.Footnote 58 Put simply, the symbolic power (or violence) of law hides the real meaning of law. Doctrinal lawyers in that sense fall victims to their own trade. This strong Bourdieusian reading exemplifies the replacement of doctrinal legal science by empirical social science as a means to explaining law. It provides a negative argument of the explanatory capacity of doctrinal law and an explicit turn to empirical inquiries of law.
Somewhat related, we have ourselves sought to develop a different framework for making IL intelligible by proposing an up-dated realist analysis of law which understands law not as a self-standing (or self-contained) normative structure but as a set of practices of knowledge and symbolic power, which eventually produces norms, including legal norms that are empirically valid.Footnote 59 Dubbed European New Legal Realism (ENLR), this approach takes seriously inter-subjectively shared ideas of law among legal agents, but it rejects that these ideas are in themselves valid – they are only held to be valid and as such they are empirically observable practices in specific social structures.
This simultaneously illustrates a subtle way in which by doing empirical work on IL it is simultaneously possible to develop an independent a posteriori argument in support of the first key claim of the replacement thesis, i.e., the negative claim about the impossibility of providing epistemological foundations for the doctrinal study of IL. More specifically, this argument takes the shape of an error theory in the sense originally developed by John Mackie in his argument for moral scepticism. Admitting the fact that there exists widespread belief in objective values, Mackie points to:
[T]he possibility of explaining, in terms of several different patterns of objectification, traces of which remain in moral language and moral concepts, how even if there were no such objective values people might have come to suppose that there are but also might persist firmly in that belief.Footnote 60
Correspondingly, as illustrated through Weber's and Ross's theory of empirical validity, ENLR claims that the empirical study of doctrinal scholarship shows how it is indeed possible to explain how, even if there is no such thing as an epistemologically sound doctrinal science of IL, legal scholars may nevertheless have come to persist firmly in that belief. And it is precisely the latter which becomes the object of inquiry for understanding law. And it is the latter – the shared collective belief – which makes it societally important.
A fourth and final example of the empirical turn in studies of IL can be found in work derived from the so-called Empirical Legal Studies (ELS), a movement which to a large extent relies on quantitative studies and in part overlaps with Law and Economics. Different from the examples of more qualitative scholarship just mentioned, ELS is where the big data revolution meets law. Generally, the claim of ELS is that results derive from testing in data. The basic assumption is that legal questions can be answered empirically.Footnote 61 But, in order to do that, so-called legal questions have to be turned into something empirically testable. In other words, legal problems have to be turned into observable patterns via a priori conceptualization. This perhaps at first glance innocent manoeuvre entails however not just a simple translation into social scientific vocabulary but rather a transformation of what is the (legal) problem in the first place.Footnote 62 Whether explicitly stated or not, ELS consist of a rejection of the hermeneutic ways of interpretation of doctrinal legal scholarship based on normative ideas and notions with a data-driven analysis which finds its explanation in more rationalistic models of interpretation (economic or other). In some instances it thereby effectively rejects doctrinal law and puts in its place an empirical quantitative explanation of law. In other instances, this kind of data-driven analysis is intended mainly as a supplement to doctrinal analysis and thus more in line with the toleration position outlined above.Footnote 63 Among Law and Economics scholars, this has caused a split between the purist economic analysis of law and a more moderate stance, arguing for the importance of legal analysis in addition to the economic study.Footnote 64
3. Conclusion
The current debate on the proper place of empirical studies of law in legal science and its relationship to more normative approaches – be it legal formalism or natural law – is, at its core, not novel. It has in various ways been discussed since the advent of modern legal science in the nineteenth century. Although Hans Kelsen's œuvre undoubtedly remains a high point in the development of a distinct legal science, not even at the time of Kelsen was it accepted as the only way of going about studying law. The success of doctrinal law more generally, notably in Europe and in countries heavily influenced by European approaches to legal research, might give the impression that doctrinal law is the science of law. This overlooks how American legal realists managed a successful counter-attack on formalism in the United States, which has had enduring effects on American legal studies: from Law and Society to Law and Economics and Empirical Legal Studies. In fact, the European idea of legal science when dealing with doctrinal law has, in the United States, largely been replaced by the more ambiguous notion of legal scholarship.
The history of international legal scholarship is no exception to these general trends in legal academia. From early on, academics of IL looked beyond formal law.Footnote 65 The second President of the Permanent Court of International Justice Max Huber (1874–1960) pioneered the use of sociology in studies of IL in his work as a professor of law at the University of Zurich before being appointed to the international bench.Footnote 66 Martti Koskenniemi likewise argues that French public international law scholarship in the period 1871–1950 had a distinct sociological dimension.Footnote 67 Another example of a sociological turn is the so-called the École de Bruxelles in law. In the United States we can observe parallel developments, for example the work of Myres S. McDougal and the Yale School of IL which mobilized sociological insights; similarly, the movement of Critical Legal Studies has multiple references to sociology and has had a significant impact on IL scholarship. The evolution of the use of political science approaches to study IL provides comparable examples of the emergence of different schools of empirical or semi-empirical assessment.
Although the fundamental problematique of the interface of doctrinal law and empirical approaches to law is therefore not new, it has undoubtedly become more pressing in light of the recent boom in empirical approaches to IL over the last couple of decades. In this article we have argued that the interface of doctrinal law and empirical approaches to law cannot be understood in partisan binaries. As suggested by our three main groupings of scholarship, as well as the differences within each ideal-typical position, there is a significant variation with regard to handling this issue. Our model nevertheless provides the conceptual space for re-describing the actual interface between doctrinal law and empirical approaches on epistemological grounds. Even when particular academic works on the topic does not fit perfectly into the taxonomy, or perhaps especially when they do not, the three categories are useful with a view to teasing out and better articulating implications of and interrelations with other work in the field. As such, the model has the following three benefits:
-
1. It provides a framework to better situate arguments about the role of empirical studies in IL;
-
2. It helps identify real epistemological stakes in order to overcome ‘trench wars’ – or worse: absence of dialogue and genuine argument;
-
3. It thus contributes to the development of a genuine basic science of law.
All three elements are important in light of the current state of debate on the proper place of empirical studies in IL. Doctrinal lawyers have often been highly defensive when empirical legal scholars have made inroads into the legal field. But more often than not the preferred strategy on both sides has been to avoid direct debate with the opposing side. Some may do this out of genuine agnosticism – simply doing their doctrinal or empirical work respectively within their specialized disciplinary domain and having no strong opinion on the character of work in the opposite camp. Others, however, actually do take a more critical stance, and yet they tend to stay silent. When in their own quarters, doctrinal and empirical legal scholars alike may renounce the approaches of the other side considering them epistemologically problematic and inadequate for explaining many current issues of IL, yet they rarely state and defend this view in open debate but rather tend to tacitly presuppose it. In this their attitude is not unlike that of Nietzsche's Zarathustra who immediately upon learning about the old saint's belief in God terminates their dialogue with a few courtesies, reserving his true opinion about such archaic beliefs for his own heart. Silently contemplating that ‘God is dead’ is, however, unlikely to produce major breakthroughs in the advancement of legal science of IL. That requires addressing head-on the ways in which doctrinal law and empirical studies might interact – and where they might not. Our framework is designed to allow precisely that critical debate.
